Improving the accuracy and efficiency of respiratory rate measurements in children using mobile devices

Abstract

The recommended method for measuring respiratory rate (RR) is counting breaths for 60 s using a timer. This method is not efficient in a busy clinical setting. There is an urgent need for a robust, low-cost method that can help front-line health care workers to measure RR quickly and accurately. Our aim was to develop a more efficient RR assessment method. RR was estimated by measuring the median time interval between breaths obtained from tapping on the touch screen of a mobile device. The estimation was continuously validated by measuring consistency (% deviation from the median) of each interval. Data from 30 subjects estimating RR from 10 standard videos with a mobile phone application were collected. A sensitivity analysis and an optimization experiment were performed to verify that a RR could be obtained in less than 60 s; that the accuracy improves when more taps are included into the calculation; and that accuracy improves when inconsistent taps are excluded. The sensitivity analysis showed that excluding inconsistent tapping and increasing the number of tap intervals improved the RR estimation. Efficiency (time to complete measurement) was significantly improved compared to traditional methods that require counting for 60 s. There was a trade-off between accuracy and efficiency. The most balanced optimization result provided a mean efficiency of 9.9 s and a normalized root mean square error of 5.6%, corresponding to 2.2 breaths/min at a respiratory rate of 40 breaths/min. The obtained 6-fold increase in mean efficiency combined with a clinically acceptable error makes this approach a viable solution for further clinical testing. The sensitivity analysis illustrating the trade-off between accuracy and efficiency will be a useful tool to define a target product profile for any novel RR estimation device.

Conflict of interest statement

Competing Interests: The authors have no patents related to the presented topic. The authors have declared that no competing interests exist.

Figures

Figure 1. Estimation of respiratory rate (RR) with a set size of 3 tap time intervals.

The median time interval is calculated for the set and the consistency C is derived. RR is reported and the measurement stopped if C is below a specified consistency threshold , otherwise tapping continues and a new set is created, until an acceptable C is obtained.

Figure 2. Screenshot of tapping screen of the RRate application.

A button records taps and an indicator displays how many taps have been performed (bottom).

Figure 3. Screenshot of feedback screen of the RRate application.

An animated (chest, shoulder and mouth) baby presents the RR. The timing of the animation can be reset with a tap. The consistency of the tap intervals is displayed on the bottom of the screen as blue dots.

Figure 4. Improvement of RR estimation accuracy with larger set sizes z (continuous curve).

The normalized root mean square error (NRMSE) of counting taps in 60 s is depicted as a dashed line; the NRMSE of the RR obtained from the median tap times of all taps in 60 s is depicted as a dotted line.

Figure 5. Plot of normalized root mean square error (NRMSE) against consistency threshold for different number of time intervals in a set.

NRMSE decreased (accuracy improved) with tighter consistency thresholds and with increasing number of intervals in a set. At lower ThC and higher z there were fewer successfully completed cases (Figure 7) that contributed to the results.

Figure 6. Plot of mean efficiency for measurement against consistency threshold ThC for the different number of time intervals in a set (z).

became greater with tighter ThC and with increasing z.

Figure 7. Completion rate CR against consistency threshold ThC for the different number of time intervals in a set (z).

Figure 8. Plot of normalized root mean square error (NRMSE) against mean efficiency for different consistency thresholds ThC and different number of time intervals in a set (z).

Increasing the accuracy (smaller NRMSE) is at the cost of increasing .

Figure 9. Distribution of normalized root mean square error (NRMSE) of the optimization results for experiments with consistency threshold (ThC = 13) and without.

Adding a ThC improves the NRMSE significantly. The horizontal lines of each box are the lower quartile, median, and upper quartile values (from bottom to top). The whiskers represent the most extreme values within 1.5 times the interquartile range from the quartile. The outlier (circle) is a value beyond the interquartile range.

Figure 10. Bland-Altman plot of the optimal parameter configuration (z = 4, ThC = 13) for the RRate application, determined by data from Phase I and II (n = 300).

The mean difference (bias) is −0.13 breaths/min and the standard deviation (SD) is 1.98 breaths/min. The number of observations is displayed as marker intensity. The dashed lines represent the 95th percentile range. The vertical lines correspond to the limits for fast breathing by age .